In the light of the relatively low number in the type IV tear category, as well as to minimize random variability, patients reporting type IV tears were combined with those reporting type II tears. By doing this, the frequency of smokers among the patients with at least a type II tear resulted in 34.8 % and differed significantly from the 23.2 % frequency among the type I tear patients (χ21 = 4.550, p = 0.033). A logistic regression model using positive smoking history as the outcome variable and including age and gender as covariates showed increasing frequencies of smokers across patients with increasing severity of tears, with all frequencies being significantly higher compared with patients with a type I tear. Similar results were obtained when combining type II with type III tear patients, or type II with type III and type IV tear patients. When considering only 131 smokers, the ANCOVA model incorporating age, gender, and duration of smoking exposure as covariates showed increasing daily average numbers of cigarettes across patients with increasing severity of tears, although the differences were only suggestive of significance (F(3.124) = 2.133, p = 0.099). The number of daily cigarettes and standard deviation per type of cuff tear are shown in Fig. 1. These differences had significant results when amalgamating type II with type III tear patients (F(2.125) = 3.222, p = 0.043), or type II with type III and type IV tear patients (F(1.126) = 4.317, p = 0.040). The ANCOVA model also indicated that adjusting by age and gender, the total number of cigarettes smoked in life differed significantly between patients with a type I tear and those with at least a type II tear (F (1.127) = 4.694, p = 0.032).